An original version of this article appeared in the Pharmaceutical Executive.
The AI revolution is shaping numerous facets of our lives. While advancements by OpenAI and Anthropic plus NVIDIA’s market performance have commanded a significant portion of media attention, the overall impact of advanced computational methods in biopharma has been largely underreported. On the front lines of drug development, a combination of AI with computational modelling and bioinformatics has become fundamental for the advancement of drug development research we’ve seen over the last decade.
Drug discovery has historically been a long, inefficient and protracted process, heavily reliant on serendipity, often tedious persistence and, good ole’-fashioned trial and error. This meant time. Lots and lots of time. Often taking over decades. And lots of money. Lots and lots of money. Sometimes hundreds of million dollars to merely identify and validate the right therapeutic targets. Then, over a billion dollars thereafter for clinical and regulatory validation. Most importantly, however, the amount of time and money spent has not always translated to improved chances of clinical and commercial success that remains a woeful <5% for many therapeutic indications.
A drastic remaking of how drugs are discovered and clinically validated is now afoot. Today, advanced computational simulations can dramatically reduce the time and cost of drug discovery and development. Most importantly, AI-guided drug discovery and development can significantly improve the odds of success particularly when combined with precision biomarker-driven companion diagnostics that pinpoint populations of patients most likely to respond to a given therapy. Quicker, cheaper and more accurate drug discovery and development can incentivize more value-based pricing while continuing to fuel the innovation engine that has dramatically improved outcomes for many diseases.
Application of advanced computational methods in drug discovery and development can create theoretical and mechanistic frameworks that streamline the process of target validation to clinical candidate selection and more precise patient identification. All of which reduce development timelines and increase the likelihood of clinical success. This is not just a technological advancement; it is a paradigm shift that promises to address unmet medical needs and improve patient outcomes on a global scale.
Two recent pieces have raised doubts about the current state of AI-assisted pharmaceutical development. Derek Lowe’s article for Science, “AI Drugs So Far,” questions the classification of AI-discovered drugs, noting that many targets were already known and not uniquely identified by AI. The Nature editorial, “AI’s potential to accelerate drug discovery needs a reality check,” highlights the promises of AI in speeding up drug development but stresses the need for independent verification and high-quality data, while calling for collaboration and transparency to maximize AI’s potential. While both articles raise valid points about the current limitations and challenges of AI in drug discovery, the seamless combination of AI with advanced computational modeling and bioinformatics approaches is effective in characterizing novel therapeutic targets and developing new first-in-class therapeutics for human diseases.
The proofs of this phenomenon across the industry are too hard to ignore.
At NImmune, for example, we create computer models, provide data, and leverage the NIMML Institute’s AI-powered TITAN-X platform to drive our precision inflammation & immunology (I&I) drug development. TITAN-X is a proprietary precision medicine platform that has shown it can accelerate the development of novel I&I therapeutics.
Building on work under NIAID Modeling Immunity for Biodefense (MIB) and the study of I&I as a massive and dynamically interacting system, the TITAN-X platform combines advanced computational modeling, bioinformatics tools and AI, including mechanistic models of CD4+ T cell activation and differentiation, multiscale models of gut mucosal immune responses, modeling pipelines with integrated machine learning tools, and advanced data analytics and visualization, plus bioinformatics pipelines, including the Modeling Metabolism (M2) module. Its computational power can inherently reduce costs, enhance clinical success rates, and identify biomarker signatures that predict treatment responses in Ulcerative Colitis and Crohn’s disease, aiding in patient selection, reducing timelines, lowering costs, de-risking product candidate and ultimately accelerating the commercialization of First-in-Class or Best-in-Class drugs. In short, TITAN-X discovers novel targets.
This integration of computational modeling, bioinformatics tools, AI-driven insights and, yes, human led medicinal chemistry, experience and ingenuity, is not just the future. It is revolutionizing our precision medicine approach to inflammatory and autoimmune disease treatment today.
In the case of NImmune, we have already utilized TITAN-X to develop an expansible pipeline of therapeutics for I&I, including two lead therapeutic assets. Omilancor is a first-in-class, oral, gut-restricted, once-daily, therapeutic undergoing Phase III clinical development for Ulcerative colitis and Phase II trials for Crohn’s disease. Omilancor’s novel mechanism of action, which was discovered and characterized by TITAN-X, creates a favorable regulatory microenvironment in the gut, decreasing the production of key inflammatory mediators and increases anti-inflammatory functions in regulatory T cells (Treg) within the site of inflammation. NIM-1324 is another first in class candidate discovered by TITAN-X. It is an oral, systemically distributed, small-molecule therapeutic candidate that is Phase II-ready for lupus and rheumatoid arthritis. Both omilancor and NIM-1324 have been developed with the help of AI and activate LANCL2, a novel therapeutic target that is responsible for modulating key cellular and molecular changes at the intersection of immunity and metabolism tied to widespread and devastating autoimmune diseases.
The market has also demonstrated clear external validation. Crucially, AbbVie acquired Landos Biopharma, a company the current NImmune leadership team founded, built from the ground up and successfully took public. Therefore, TITAN-X discoveries have now been validated not merely by the clinical success of omilancor and NIM-1324, targeting LANCL2, but also commercially through AbbVie’s acquisition of Landos Biopharma.
AbbVie’s focus on Landos’ lead candidate NX-13 (now called ABBV-113 following the acquisition by AbbVie), targeting NLRX1, characterized leveraging TITAN-X technology, reflects an understanding that there is demand for new options, specifically for cutting-edge computational drug development. This transaction gives us tremendous confidence about the demand for advanced AI in drug development. Other market signals include Merck’s $10.8 billion acquisition of Prometheus, which also included the advanced computational engine that supported their drug development, another novel mechanism of action, and a lead asset with similar efficacy to omilancor in UC. In a phase 2, proof-of-concept, double-blind, randomized, placebo-controlled trial, oral omilancor induced clinical remission in 30.4% of patients with active UC (78% with baseline Mayo endoscopic subscore [MES] of 3) relative to 3.7% of the placebo arm (Δ = 26.7, P = 0.01). In a similar Phase 2 UC trial, PRA-023, an injected anti-TL1A drug candidate, induced clinical remission in 26.5% of patients with active UC (68% with baseline MES of 3) relative to 1.5% of the placebo group (Δ = 25).
As we continue to innovate and harness the power of AI in I&I research and discovery, we are encouraged by the clear successes and validations in the market. The influx of capital and talent into AI, coupled with its exponential development, has already yielded substantial returns for us at NImmune, particularly through our TITAN-X platform. The growing interest and investment in AI-driven drug discovery highlight its transformative potential. Importantly, AI is a tool to make drug discovery cheaper, faster, and better. AI is not a replacement for human ingenuity.
At NImmune BioPharma, we will continue to leverage AI and advanced computational modeling to push the boundaries of what is possible in I&I drug development. We remain committed to continuing to grow the TITAN-X platform‘s capabilities to address a broader range of I&I diseases, including lupus, psoriasis, atopic dermatitis, asthma, multiple sclerosis, diabetes, and Alzheimer disease, and continually refine our models, clinical datasets and predictive algorithms to enhance patient stratification and therapeutic outcomes.
I am excited about the future of AI-enabled advancements that will shape the next generation of precision medicines, improving the lives of patients worldwide.
Dr. Josep Bassaganya-Riera is the Founder, Executive Chairman, President and CEO of NImmune Biopharma, Inc. He also serves as President and Founding Director of the NIMML Institute, a 501 (c)(3) public charity foundation.